Traction and tractor_performance

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/237106038
Traction and Tractor Performance
Article · March 2012
CITATIONS
50
READS
1,452
2 authors, including:
Some of the authors of this publication are also working on these related projects:
Biomass Logistics View project
Performance measurement of three-point mounted implement guidance systems View project
Robert D Grisso
Virginia Polytechnic Institute and State University
169 PUBLICATIONS 1,145 CITATIONS
SEE PROFILE
All content following this page was uploaded by Robert D Grisso on 08 October 2014.
The user has requested enhancement of the downloaded file.

ASAE DISTINGUISHED LECTURE SERIES
Tractor Design No. 27
Traction and Tractor
Performance
Frank M. Zoz and Robert D. Grisso
These lectures have been developed to provide
in-depth resource information for engineers
in the agricultural industry.

Copyright © 2003 by the American Society of Agricultural Engineers
All Rights Reserved
Manufactured in the United States of America
This lecture may not be reproduced in whole or in part by any means
(with the exception of short quotes for the purpose of review)
without the permission of the publisher
For information, contact:
ASAE, 2950 Niles Rd., St. Joseph, MI 49085-9659 USA.
Phone: 269 429 0300 Fax: 269 429 3852 www.asae.org
ASAE Publication Number 913C0403
The American Society of Agricultural Engineers is not responsible for statements
and opinions advanced in its meetings or printed in its publications.
They represent the views of the individual to whom they are credited
and are not binding on the Society as a whole.

Traction and Tractor
Performance
Frank M. Zoz
Retired
John Deere Product Engineering Center
Waterloo, Iowa, USA
Robert D. Grisso
Professor
Virginia Tech
Blacksburg, Virginia, USA
For presentation at the 2003 Agricultural Equipment Technology Conference
Louisville, Kentucky, USA
9-11 February 2003
Published by
ASAE – the Society for engineering in agricultural, food, and biological systems
2950 Niles Road, St. Joseph, MI 49085-9659 USA
The Lecture Series has been developed by the Power and Machinery Division Tractor
Committee (PM-47) of ASAE to provide in-depth design resource information for
engineers in the agricultural industry. Topics shall be related to the power plant, power
train, hydraulic system, and chassis components such as operator environment, tires, and
electrical equipment for agricultural or industrial tractors or self-propelled agricultural
equipment.
ASAE is grateful to Deere & Co for sponsoring the ASAE Distinguished Lecture Series.

Table of Contents
INTRODUCTION .................................................................................................................5
TRACTION MECHANICS ....................................................................................................5
Solid Wheel on a Hard Surface ...................................................................................5
Soft Wheel on a Hard Surface .....................................................................................6
Deformable Wheel on a Soft Surface ..........................................................................7
Belt Drives ................................................................................................................8
TRACTION PARAMETERS .................................................................................................9
Travel Reduction Ratio (TRR) ....................................................................................9
Net Traction Ratio (NTR) .........................................................................................10
Tractive Efficiency (TE) ..........................................................................................11
Gross Traction Ratio (GTR) .....................................................................................11
Motion Resistance Ratio (MRR) ...............................................................................11
TRACTION DATA ANALYSIS ..........................................................................................12
Pull Slip and NTR Slip .............................................................................................12
Tractive Efficiency ..................................................................................................14
Regression Analysis .................................................................................................17
TRACTION TESTING ........................................................................................................19
Single-Wheel Testing ...............................................................................................20
Using Tractors to Test Tires .....................................................................................21
Speed Effects ..........................................................................................................23
TRACTION PERFORMANCE ............................................................................................24
Effects of Soil ..........................................................................................................24
Effects of Tire Pressure ............................................................................................25
Effects of Tire Size ..................................................................................................25
Effects of Load on Tire ............................................................................................26
Belt and Tire Comparisons .......................................................................................26
SOIL, TIRE, AND TRACTION EQUATIONS .....................................................................30
TRACTOR PERFORMANCE ..............................................................................................33
Tractor Performance Spreadsheet ..............................................................................35
Estimated Drawbar Power ........................................................................................38
OPTIMIZING TRACTOR DRAWBAR PERFORMANCE ....................................................39
Tires .......................................................................................................................39
Ballasting ................................................................................................................39
Ballasting Sensitivity ...............................................................................................43
Ballasting Limitations ..............................................................................................44
Ballast Optimization in the Tractor Performance Spreadsheet .....................................44
CONCLUSIONS .................................................................................................................45
ACKNOWLEDGEMENTS ..................................................................................................45
REFERENCES ....................................................................................................................45

5
Traction and Tractor Performance
Frank M. Zoz, P.E.
Retired, John Deere Product Engineering Center, Waterloo, Iowa
Robert D. Grisso, P.E.
Professor, Biological Systems Engineering,
Virginia Tech, Blacksburg, Virginia
The primary purpose of agricultural tractors, especially those in the middle to high power ranges, is to
perform drawbar work. The value of a tractor is measured by the amount of work accomplished relative to
the cost incurred in getting the work done. Drawbar work is defined by pull and travel speed. Therefore, the
ideal tractor converts all the energy from the fuel into useful work at the drawbar. In practice, most of the
potential energy is lost in the conversion of chemical energy to mechanical energy, along with losses from
the engine through the drivetrain and finally through the tractive device. Research shows that about 20% to
55% of the available tractor energy is wasted at the tractive device/soil interface. This energy wears the
tires and compacts the soil to a degree that may cause detrimental crop production (Burt et al., 1982).
Efficient operation of farm tractors includes: (1) maximizing the fuel efficiency of the engine and
drivetrain, (2) maximizing the tractive advantage of the traction devices, and (3) selecting an optimum
travel speed for a given tractor-implement system.
Throughout the years, official tractor performance drawbar tests have been conducted on hard surfaces
and in recent years (30+ years) on concrete. While this provides a valid comparison between tractors, the
data does not provide much information about performance under field conditions. The primary difference
between official tests and field conditions is the performance of the tires or other tractive devices.
The understanding and prediction of tractor performance has been a major goal of many researchers.
Tractor performance is influenced by traction elements, soil conditions, implement type, and tractor
configuration (Brixius, 1987). It is necessary to understand traction performance to predict tractor
performance in the field. Traction equations provide a basis for predicting tractor performance when
combined with basic information from official tractor tests.
Computer models allow researchers and designers to investigate many problems related to tractor
performance under a wide range of conditions with the goal to improve tractor design, to optimize tractor
operational parameters, and to improve the tractor/implement match. Relative importance of these factors
affecting field performance of a tractor can be achieved without expensive field-testing. For teachers,
models enhance the student's ability to comprehend and compare various parameters that influence tractor
performance. These models can also assist tractor operators to improve (fine-tune) and optimize their
tractors' setup to match operating conditions.
1. Traction Mechanics
Solid Wheel on a Hard Surface
An understanding of traction mechanics is fundamental to understanding differences between tractive
performance and tractor performance. The basic forces involved in a powered wheel are shown in figure 1
for the simple case of a solid wheel on a hard surface. The torque input (T) develops a gross traction (GT)
acting at the contact surface. Part of the gross traction is required to overcome motion resistance (MR),
which is the resistance to the motion of the wheel, including internal and external forces. The remainder is
equal to the net traction (NT) that the wheel develops, given by NT = GT - MR.

6
GT MR
slrrr
NT
T
W
Va
Wd
rt
ϖϖ
W = Weight, static
Wd = Weight, dynamic
slr = Loaded radius, static
rr = Rolling radius
rt = Torque radius
Vt = Velocity, theoretical
Va = Velocity, actual
T = Axle torque
GT
NT
MR
GT MR
slrrr
NT
T
W
Va
Wd
rt
ϖϖ
W
Wd
slr
rr
rt
Vt
Va
T
GT = Gross traction (theoretical pull)
NT = Net traction (actual pull)
MR = Motion resistance
Figure 1. Basic wheel forces for a solid wheel on a hard surface.
Soft Wheel on a Hard Surface
A soft wheel on a hard surface (fig. 2) is much the same as a solid wheel except that it becomes more
obvious that the vertical reaction force (Wd) is not directly under the axle centerline but is offset by a
distance designated "eh." This offset is necessary for static equilibrium. The amount of the offset is a
function of the motion resistance and is given by:
( )( )
W d
MRslr
eh = (1)
Three radii are shown in figures 1 and 2: "slr" is the static loaded radius, defined as the distance from
the axle centerline to a hard surface; "rr" is the rolling radius, used for speed calculations. Rolling radius is
derived from the rolling circumference (usually measured prior to a test but also included in tire
manufacturer data tables). Static loaded radius and rolling radius are close but not equal. For a properly
inflated agricultural tire, the rolling radius is about 6% greater than the static loaded radius. Rolling radius
should be used for speed calculations. Static loaded radius is more appropriate to use for force or moment
calculations. Both can be affected by the softness of the soil surface and are usually determined on a hard
surface.
The third radius shown in figure 1 is called the torque radius (rt). This is the effective radius where the
gross traction (GT) and motion resistance (MR) forces act. It cannot be measured directly but can be
determined by back calculating using energy calculations. This is explained in detail in section 2 of this
paper under "Gross Traction Ratio."

7
MR
slrrr
NT
T
W
Va
Wd
GT
eh
rt
ϖϖ
W = Weight, static
rr = Rolling radius
rt = Torque radius
T = Axle torque
MR
slrrr
NT
T
W
Va
Wd
GT
eh
rt
ϖϖ
W
Wd
slr
rr
rt
Vt
Va
T
GT
NT
MR
Figure 2. Basic wheel forces for a soft wheel on a hard surface.
Deformable Wheel on a Soft Surface
In the real world, both the wheel and the (soil) surface are deformable and result in the forces and
moments shown in figure 3. The result is both a vertical and a horizontal offset, designated "ev" and "eh,"
respectively. The amount of the offsets depends on the motion resistance force (MR), the tire loaded radius
(slr), and the vertical force resultant (Wd):
( )( )
W d
MRev-slr
eh = (2)
slrrr
NT
T
W
Va
GT
Wd
eh
MR
ev
Ground LineGround Line
rt
ϖϖ
W = Weight, static
rr = Rolling radius
rt = Torque radius
T = Axle torque
slrrr
NT
T
W
Va
GT
Wd
eh
MR
ev
Ground LineGround Line
rt
ϖϖ
W
Wd
slr
rr
rt
Vt
Va
T
GT
NT
MR
Figure 3. Deformable wheel on a soft surface.

8
Belt Drives
The mechanics of the belt drive mechanism (fig. 4) is similar to the wheel in many respects; but the
distribution of the load is dependent on vehicle parameters. The location of the dynamic load resultant, "eh"
(dynamic balance ratio; Corcoran and Gove, 1985), depends on the static distribution, the design of the
suspension mechanism supporting the bogie wheels, and vehicle weight transfer characteristics.
MR
slrrr
NT T
W1
Va
Wd
GT
W2 W3 W4
W5
Ground Line
Dh
rt
ϖϖ
T = Axle torque
W = Weight, static
rr = Rolling radius
rt = Torque radius
Vt
Va
T
GT
NT
MR
W
Wd
slr
rr
rt
eh
Figure 4. Belt drive.
In general, the best tractive performance and the most uniform ground pressure on a belt drive can be
obtained with a dynamic balance ratio of near 50%. Corcoran and Gove (1985) defined dynamic balance
ratio as the ratio of the location of the vertical component of the dynamic load (external loads and tractor
weight) from the front of the track divided by the track base length. Unlike a tire, where only the total
dynamic weight (Wd) must be considered during a traction test, the dynamic balance of a track mechanism
must be considered either in a belt test or a full vehicle test. The dynamic balance ratio obtained depends
not only on tractor dimensional parameters and the location and angle of the line of draft but also on the
magnitude of the drawbar pull.
Figure 5 shows the effect of draft angle and vehicle traction ratio on the dynamic balance for a tractor
with 60% of the static weight at the front. A generic wheel/belt tractor is shown for simplicity, but the
weight transfer mechanics are the same for belted and wheel tractors. Note that it only takes a 5° draft angle
to give a 50% dynamic balance ratio at a typical vehicle traction ratio of 0.40. Figure 5 is for implements
hitched to the drawbar. Even higher weight transfer, and hence higher front weight requirements, may
result from the use of three-point hitch equipment.

9
5
0
10
20
40
42
44
46
48
50
52
54
56
58
60
0 0.1 0.2 0.3 0.4 0.5 0.6
Vehicle Traction Ratio
DynamicBalance(%Front)
Draft Angle
With
Dh/Wb = 0.20
Dl/Wb = 0.30
WbDl
Draft
Angle
Dh
0.60
0.58
0.56
0.54
0.52
0.50
0.48
0.46
0.44
0.42
0.40
DynamicBalanceRatio
5
0
10
20
40
42
44
46
48
50
52
54
56
58
60
0 0.1 0.2 0.3 0.4 0.5 0.6
With
Dh/Wb = 0.20
Dl/Wb = 0.30
WbDl
Draft
Angle
Dh
0.60
0.58
0.56
0.54
0.52
0.50
0.48
0.46
0.44
0.42
0.40
Figure 5. Tractor/belt drive dynamic weight distribution (when starting with 60% static front weight).
2. Traction Parameters
Five dimensionless parameters are used to describe tractive performance:
• Travel reduction ratio (TRR), commonly called "slip" and expressed in percent.
• Net traction ratio (NTR), sometimes called pull/weight ratio.
• Tractive efficiency (TE), usually thought of as percent but used as a ratio in this paper.
• Gross traction ratio (GTR).
• Motion resistance ratio (MRR).
The traction parameters involving forces are all normalized by dividing by Wd, the dynamic force
reaction supporting the wheel or traction device. Wd includes static axle weight and any weight transfer
that might take place during the testing process, i.e., the total reaction force. Dividing by Wd allows
comparisons between tires and other tractive devices of different sizes and weights, and provides a
dimensionless parameter for traction comparisons.
Note that all the traction parameters are normally presented as ratios except travel reduction and tractive
efficiency, which are commonly expressed as percentages. Working with traction data is easier if all
parameters are presented as ratios. It will become more obvious later, but remember that the above
parameters apply to a traction device and not necessarily to a vehicle.
Travel Reduction Ratio (TRR)
Vt
Va
1
VelocitylTheoretica
VelocityActual
1TRR −=−= (3)
Travel reduction has traditionally been called "slip" or "% slip," but technically this is incorrect. Slip
occurs between surfaces. Travel reduction is a reduction in distance traveled and/or speed that occurs
because of:
• Flexing of the tractive device

10
• Slip between the surfaces (rubber and concrete, for example)
• Shear within the soil.
From a power efficiency standpoint, travel reduction is a power loss caused by a loss in travel speed or
distance traveled. Slip (travel reduction) occurs any time a wheel or traction device develops pull (net
traction) (Brixius and Wismer, 1978).
Zero travel reduction can be defined using any of four methods (ASAE Standards, 2001b):
1. A self-propelled (zero net traction) condition on a non-deforming surface (recommended for rolling
circumference data, as in published tire data).
2. A self-propelled (zero net traction) condition on the test surface.
3. A towed (zero gross traction, i.e., zero torque) condition on a non-deforming surface.
4. A towed (zero gross traction) condition on the test surface.
There are arguments for using any of the above methods for a particular traction test. In any case, the
zero condition used to define the rolling radius should always be stated. The most common zero condition
is use of the self-propelled condition on the test surface (method 2). However, tire data are usually given
for a non-deforming surface (method 1). The difference in measured rolling radii between a non-deforming
(hard) surface and a test surface is small under normal agricultural soil conditions (dry and/or untilled soil)
and thus makes little difference in the final results. In any case, errors of defining "zero slip" do not affect
the final tractive efficiency results, as travel reduction does not enter directly into the equation. It only
affects the results where the losses are assigned, that is, either to travel reduction or motion resistance. This
will be discussed in more detail in section 3 of this paper.
The authors' preference is to use a self-propelled condition (zero net traction) on a hard surface, and this
method is used throughout this paper. This method provides a repeatable test condition, results that should
agree closely to published tire data, and data that can be replicated at other locations and test conditions. It
is also easy to imagine a case of very soft soil where the zero condition may result in an apparent 100%
slip, i.e., the vehicle gets stuck, while being assigned zero travel reduction.
The rolling radius (rr) measured under one of the above methods is used to calculate the theoretical
speed (Vt) of the wheel or tractive device:
Vt (m/s) = ϖ (rpm) × rr × 2π/60 (4)
The actual forward velocity (Va) of the vehicle or wheel is usually measured directly using a fifth wheel
or radar device. Both Vt and Va must use the same units of measurement.
Net Traction Ratio (NTR)
W d
NT
ForceReactionDynamic
TractionNet
NTR == (5)
The net traction ratio is sometimes referred to as pull/weight, P/W, dynamic traction ratio, or coefficient
of traction. Most of these terms actually refer to a complete vehicle rather than to a simple traction device.
The dynamic reaction force or dynamic weight (Wd) includes the effects of ballast and any weight transfer
that may occur in the testing process. If a complete vehicle is used for the traction device testing, the weight
may include front to rear transfer due to horizontal pull, and any transfer due to implement or load unit
draft angle. The net traction force (NT) must be the force component in the direction of travel and
perpendicular to the reaction force (Wd).
As stated, the above equation applies to a tractive device and not to a comp lete vehicle. For a total
vehicle (tractor), the equivalent to net traction ratio (NTR) is vehicle traction ratio (VTR), which is the
vehicles' drawbar pull divided by the total vehicle dynamic weight. This will be covered in more detail in
section 7 of this paper.

11
Tractive Efficiency (TE)












===
×
==
Vt
Va
GTR
NTR
Vt
Va
W d
GT
W d
NT
Vt
Va
GT
NT
PowerAxle
VaNT
PowerInput
PowerOutput
(ratio)TE
(6)
Tractive "inefficiency" is caused by both velocity losses and pull losses. The loss in travel speed is
commonly referred to as "slip," although it is more accurately refered to as "travel reduction." Travel
reduction is the result of the theoretical travel speed (Vt) not being entirely converted to forward progress
(Va) due to losses within the soil, between the soil surface and the traction device, and within the traction
device (hysteresis, and tire windup or belt slippage). Travel reduction losses are visible, that is, the operator
can see it happening. The other component of tractive "inefficiency," which is less visible and often
overlooked, is a loss of pull (net traction) when motion resistance reduces the amount of gross traction that
is converted to useful output (net traction). This is part of what happens when a tractor is overballasted.
Travel reduction is reduced, but motion resistance is increased. Motion resistance losses are especially
relevant to belts, as internal losses within the belt drive mechanism, rollers, and bending of the belt are
normally greater than those within a tire. On soft soils, the internal losses of belts are generally
compensated for by lower external motion resistance than that of tires.
Gross Traction Ratio (GTR)
W drt
T
W d
GT
GTR
×
== (7)
Gross traction (GT) is sometimes referred to as rim pull, design drawbar pull, or theoretical pull. It is
the axle torque input converted to a pull force. It is the pull you would develop if there were no motion
resistance loss.
The gross traction ratio (GTR) is the least understood of the traction parameters. Gross traction (GT)
itself cannot be measured directly and is usually calculated from the axle torque and radius of the wheel or
tractive device. The problem is that the correct radius to use is not well defined or directly measurable.
There is no general agreement among traction researchers as to what radius to use, and an alternate method
of calculating gross traction ratio is preferred using energy or power considerations.
From equations 6 and 3:
( )TRR1
TE
NTR
Vt
Va
TE
NTR
GTR −





=











= (8)
Having thus determined the gross traction ratio (GTR), since
W drt
T
W d
GT
GTR
×
== ,
the effective torque radius (rt) can be calculated (although it is only of academic interest at this point to
know its value) by:
( )( )W dGTR
T
rt = (9)
Motion Resistance Ratio (MRR)
MRR = GTR - NTR (10)
The motion resistance ratio (MRR), sometimes referred to as rolling resistance, includes internal losses
within the tractive device (for example, losses within a belt drive or a tire) and soil forces. All "force"
losses beyond where the torque is measured are included in motion resistance, for example, gear losses if
the torque is not measured directly at the input to the tractive device. An example of this might be use of
input drive shaft torque when testing tires using a mechanical front wheel drive mechanism. Rolling losses

12
of bogie wheels of a belt drive mechanism would be another example, as well as the torque required to
overcome the bending of a belt.
3. Traction Data Analysis
Traction data are usually normalized to create dimensionless ratios by dividing the data by the total
vertical ground reaction force (Wd). Aside from making it easier to compare tractive devices of differing
sizes, the traction performance ratios make it easy to plot and compare the traction parameters for a single
tractive device.
Pull Slip and NTR Slip
The most basic plot of traction data is to simply compare the pull to the travel reduction (slip). This is
often the only data considered for a device where power consumption may not be a primary consideration
and/or where traversing a certain terrain may be the total objective. There is also a question of which is the
independent variable, pull or slip? Traction data has historically considered slip (travel reduction) to be the
independent variable, that is, developing pull depends on travel reduction (slip). However, this is not
unanimous, and there are reasons to believe that pull (NTR) is the independent variable and that "slip
happens." Most of the data shown in this paper will consider NTR (or VTR for a complete vehicle) to be
the independent variable.
Figure 6 is an example of a pull-slip (travel reduction) plot showing a Goodyear belt on a John Deere
tractor. Note that the pull rises steeply and then begins to level off as travel reduction increases. It will
reach some maximum point and likely drop off as slip increases further. Most data shown in this paper will
not exceed 40% travel reduction because peak tractive efficiency will be shown to occur at a lower level of
pull. In addition, consider that the pull that might be developed also depends on the weight on the traction
device, which is shown to be 11791 kg (26000 lbs).
0
20
40
60
80
100
120
-0.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Travel Reduction Ratio
DrawbarPull,kN
0
20
40
60
80
100
120
0
20
40
60
80
100
120
-0.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
DrawbarPull,kN
Figure 6. A typical tractor pull-slip (travel reduction) curve (John Deere 8400T, total mass = 11791 kg, belt = Goodyear
400 mm), surface soil plowed and settled.
Another example of a pull-slip (travel reduction) plot is shown in figure 7 from Nebraska Tractor Test
Laboratory (NTTL) data (for a complete vehicle). It also shows why dividing by the dynamic weight on the
tractive device to create dimensionless ratios makes the comparisons more meaningful. A casual look at the
travel reduction plots makes it appear that the John Deere 8400 ballasted tractor is outperforming the
others, which in this case included both belt and rubber-tired tractors, both ballasted and unballasted.

13
JD 8400 MFWD Ballasted
JD 8400 MFWD Unballasted
CAT 45 Ballasted
0
5000
10000
15000
20000
25000
30000
0 2 4 6 8 10 12 14 16
Travel Reduction (%)
DrawbarPull(lbs)
CAT 45 Unballasted
JD 8400 MFWD Ballasted
JD 8400 MFWD Unballasted
CAT 45 Ballasted
0
5000
10000
15000
20000
25000
30000
0 2 4 6 8 10 12 14 16
CAT 45 Unballasted
Figure 7. Nebraska test drawbar pull comparison (ballasted and unballasted).
However, dividing the pull by the weight of the vehicles, in this case the total weight on the tractive
devices produces a different picture (fig. 8).
0.00
0.20
0.40
0.60
0.80
1.00
1.20
0 2 4 6 8 10 12 14 16
VehicleTractionRatio,VTR(pull/weight)
CAT 45
JD 8400 MFWD
0.00
0.20
0.40
0.60
0.80
1.00
1.20
0 2 4 6 8 10 12 14 16
CAT 45
JD 8400 MFWD
Figure 8. Nebraska test vehicle traction ratio comparison (ballasted and unballasted).
And there is still the question of which is the independent variable: travel reduction (slip), or pull (NTR
or VTR)? This question cannot be answered here, as two definite groups support these two viewpoints, and
each has supporting reasons. However, it is the authors' opinion that slip (travel reduction) happens because
of a pull (NTR or VTR) being applied (fig. 9). This assumption will be used for the remainder of this paper.

14
0
2
4
6
8
10
12
14
16
0.00 0.20 0.40 0.60 0.80 1.00 1.20
TravelReduction(%)
Vehicle Traction Ratio, VTR (pull/weight)
“Slip Happens!” That is, Slip depends upon pull.“Slip Happens!” That is, Slip depends upon pull.
The pull related term, NTR, GTR, VTR shouldThe pull related term, NTR, GTR, VTR should
be the independent variablebe the independent variable
Cat 45 Unballasted
Cat 45 Ballasted
0
2
4
6
8
10
12
14
16
0.00 0.20 0.40 0.60 0.80 1.00 1.20
"Slip happens!" That is, slip depends on pull.
The pull-related term (NTR, GTR, VTR) should
be the independent variable
Cat 45 Unballasted
Cat 45 Ballasted
Figure 9. Nebraska test performance comparison with VTR as independent variable.
Tractive Efficiency
From a tractor drawbar power standpoint, tractive efficiency (TE) is the most important of the traction
parameters. Figure 10 is a generalized plot of the traction relationships using the Brixius (1987) traction
equations and net traction ratio (NTR) as the independent variable. For a properly ballasted and inflated
agricultural tire, tractive efficiency (TE) tends to maximize at a net traction ratio (NTR) of approximately
0.40. This was also recognized by Dwyer (1984). Motion resistance ratio (MRR) tends to be a linear
function of either slip or NTR unless slip sinkage becomes a factor.
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
TractionPerformanceRatio
Net Traction Ratio
(NTR)
Travel Reduction Ratio)
Tractive Efficiency Ratio (TE)
(GTR - NTR)
Motion Resistance Ratio
(1 - Va/Vt)
(GTR)
Gross Traction Ratio
Figure 10. Generalized traction relationship based on the Brixius (1987) traction equation.
There are travel and force losses that make up tractive "inefficiency" that do not have official
terminology. The simplest way to understand them is to look at them as a "velocity ratio" and a "pull ratio."
In equation form:
RatioVelocityRatioPull
Vt
Va
GTR
NTR
TE ×=











= (11)
The velocity ratio is shown as a function of net traction ratio (NTR) in figure 11. At zero NTR (zero
pull) the actual velocity (Va) is about equal to the theoretical velocity (Vt), depending somewhat on the
definition of "zero" slip (ASAE Standards, 2001b), and the velocity ratio is near unity. As net traction (pull)

15
increases, travel reduction (slip) increases, and the velocity ratio decreases. For a given traction device,
velocity ratio losses depend on the characteristic shape of the pull-travel reduction curve.
Va/ Vt
Travel Reduction Ratio = (1-Va/Vt)
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Va = Actual Velocity
Vt = Theoretical Velocity
T E
N T R
G T R
V a
V t
=












0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Va = Actual velocity
Vt = Theoretical velocity
T E
N T R
G T R
V a
V t
=












Velocity Ratio (Va/Vt)
(1 - Va/Vt)
Figure 11. Velocity losses in traction.
Pull ratio is shown as a function of NTR in figure 12. At zero net traction (pull), the ratio of net traction
ratio (NTR) to gross traction ratio (GTR) approaches zero; the difference between GTR and NTR is the
motion resistance ratio (MRR), which is in the range of 0.05 to 0.15. Due to motion resistance, the net
traction ratio can never equal the gross traction ratio, and the pull ratio approaches but never reaches unity.
- NTR
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
(GTR -NTR)
Pull Ratio (NTR / GTR)
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
T E
N T R
G T R
V a
V t
=











T E
N T R
G T R
V a
V t
=












Figure 12. Pull losses in traction.
Tractive efficiency is defined as output power / input power. It can also be expressed as the product of
pull ratio and velocity ratio (eq. 11). Figure 13 shows how velocity ratio and pull ratio combine for overall
tractive efficiency. The overall tractive efficiency cannot be greater than either the pull ratio or the velocity
ratio, and thus it reaches a maximum value at NTR of between about 0.3 and 0.4 with radial-ply tires. It
will be shown later that a similar range exists for belts.

16
Tractive Efficiency
Ratio
Velocity Ratio = Va / Vt
Travel Reduction Ratio = (1 -Va/Vt)
Ratio = GTR - NTR
Maximum Efficiency Point
Pull Ratio = NTR / GTR
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
T E
N T R
G T R
V a
V t
=












Tractive Efficiency
Ratio
Velocity Ratio
(Va/Vt)
(GTR - NTR)
Maximum Efficiency Point
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
T E
N T R
G T R
V a
V t
=












(1 - Va/Vt)
Figure 13. Overall tractive efficiency with velocity and pull losses.
Pull ratio, velocity ratio, and tractive efficiency are shown for real data in figure 14. Data is for radial-
ply tires in medium (tilled) tractive conditions. The curves are the result of regression analysis of the field
test data. Both velocity (travel reduction) and pull (motion resistance) losses contribute to the overall
tractive efficiency.
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Gross Traction
Ratio
Velocity Ratio
(Va/Vt)
Tractive Efficiency
Figure 14. Traction data with regression and loss curves. Tire = Goodyear 20.8R42 DT/DT710 dual. Surface = Lon's tilled
(seeded).
The same power loss data can be viewed in a travel reduction (slip) based plot (fig. 15), but it is not so
easy or intuitive to relate the losses to travel reduction. This is probably because we are looking at losses as
a function of one of the losses itself. This again supports NTR as being the independent variable.

17
Tractive Efficiency Net Traction Ratio
Pull Ratio = NTR/GTR
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0 10 20 30 40 50
Velocity Ratio =Va/Vt
T E
N T R
G T R
V a
V t
=












Motion Resistance Ratio = GTR - NTR
Tractive Efficiency Net Traction Ratio
Pull Ratio (NTR/GTR)
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0 10 20 30 40 50
Velocity Ratio (Va/Vt)
T E
N T R
G T R
V a
V t
=












Motion Resistance Ratio (GTR - NTR)
Figure 15. Traction losses with travel reduction (slip) as independent variable.
Regression Analysis
Typical traction data plots are shown in figures 16 and 17. Figure 16 shows an example of the
traditional plot with travel reduction as the independent variable. Each point on the plot represents the
average of 1.5 seconds of data sampled at approximately 50 Hz. Considerable data scatter occurs,
especially at the lower values of travel reduction. Most of the data scatter can probably be attributed to the
difficulty in measuring actual travel speed (Va). This is normally done with a fifth wheel or a radar device.
Variations in Va affect the travel reduction calculation, especially at values near zero. Thus, the apparent
vertical scatter of the NTR (pull) data is actually a horizontal scatter of the travel reduction data.
Variations in the measurement of actual travel speed (Va) also affect the calculated tractive efficiency,
which shows a similar scatter at the lower travel reduction values. Note that while the motion resistance
ratio (MRR) is also a calculated value, it does not show the same increase in scatter at the low travel
reduction values. This is because motion resistance is calculated from gross traction and net traction, which
do not show a variation with travel reduction. In addition, as motion resistance ratio (MRR) is nearly
constant with travel reduction, any variations caused by variations in the slip calculation will be hidden as
the points simply move sideways.
The resulting scatter of the data makes it difficult to determine significant values to use to compare
performance of two (or more) tractive devices. This is one reason that regression analysis of the data is
commonly used. The method used by the authors was developed with the assistance of Dr. Shrini
Upadhyaya of the University of California, Davis (Upadhyaya et al., 1988). His Quick Basic program for
Macintosh was converted to run in Excel Visual Basic. The original method included a regression of GTR
and NTR as a function of slip (travel reduction). It was modified in the conversion to Excel Visual Basic to
regress NTR and MRR. The method determines a regression curve for NTR and MRR as a function of
travel reduction (slip):
NTR = DA × [1 - exp(-C0 × TRR × 100)] (12)
MRR = TA + AB × TRR (13)
Following determination of the regression coefficients, GTR and TE can be calculated as:
GTR = NTR + MRR
TE = NTR × (1 - TRR) / GTR
In the following examples (figs. 16 through 19), each data point represents the average for 1.5 seconds.
This example shows the merging of three replications of the test run. Also shown as a smooth curve are the

18
results of a regression of the NTR and MRR data. These are examples only, and all subsequent
comparisons are made using the regression coefficients.
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
-0.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Tractive Efficiency
Net Traction Ratio
Figure 16. Traditional slip plot with experimental data. Tire = 20.8R42 dual. Surface = Lon's tilled (seeded).
1.00
-0.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
NTR = 0.664(1-exp-0.098(TRR))
MRR=0.090+0.001(TRR))
MotionResistanceRatio
Tractive Efficiency
NetTractionRatio
Figure 17. Traditional slip plot of traction data with regression curves. Tire = 20.8R42 dual.Surface=Lon'stilled
(seeded).

19
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.000.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Net Traction Ratio, NTR
Tractive Efficiency
Figure 18. Experimental data with net traction ratio as independent variable. Tire = Goodyear 20.8R42 DT/DT710 dual.
Surface = Lon's tilled (seeded).
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Net Traction Ratio, NTR
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
NTR = 0.664(1-exp-0.098(TRR))
MR = 0.090 + 0.001(TRR))
Tractive Efficiency
Figure 19. Experimental data and regression curves with net traction ratio as independent variable. Tire = Goodyear
20.8R42 DT/DT710 dual. Surface = Lon's tilled (seeded).
4. Traction Testing
Traction testing involves operating a traction device (wheel or belt) in the soil and making
measurements of its performance. Four dynamic (on the go) measurements are required:
• Input torque (T)
• Input speed (ω)
• Output force (NT)
• Output speed (Va)
Additionally, the dynamic weight (ground reaction force) must be known, measured, or calculated. This
will usually depend on the design of the traction test device. If a single-wheel tester is used, it can be
designed such that the dynamic weight reaction force is equal to that measured statically.

20
Single-Wheel Testing
The simplest device for a traction test of a wheeled device requires supporting the moving wheel,
applying the required torque, and measuring the developed force (net traction). However, there are various
ways this can be accomplished, with varying levels of complexity. Some devices can operate only in soil
bins, i.e., bring the soil to the device, while others are operated in the field. In some cases, testing is done
using complete vehicles, with the tractive device being the drive wheels or tracks. The most basic traction
testing device is shown in figure 20.
GT MR
slrrr
NT
T
W
Va
ωω
Wd
rt
GT MR
slrrr
NT
T
W
Va
ωω
Wd
With a single arm, torque (T) results
in a change in vertical reaction,
which must then be measured
during the test.
rt
Figure 20. Simplest form of single-wheel tester.
With the single-link device (fig. 20), a change in input torque (T) results in a change in vertical force
reaction (Wd), which then must be measured dynamically during the test (normally set and measured
statically). Figure 21 shows a modification using two parallel links; this eliminates the weight transfer
effect but may result in more difficult measurement of NT and T. Actually, there are two alternatives for
measuring torque: directly measuring the input torque, or determining torque from the measurement of NT.
Net traction is the vector sum of the two reaction forces; the input torque can be determined by the
difference in the two reaction forces multiplied by the distance between the links.

21
GT MR
slrrr
T
W
Va
ωω
Wd
NT/2
With parallel arms, there is
no change in vertical reaction
as torque is applied.
NT/2
Net traction can be calculated
as the sum of the reaction forces.
Torque can be calculated as the
difference in the forces multiplied
by the distance between the arms.
rt
GT MR
slrrr
T
W
Va
ωω
Wd
NT/2
NT/2
rt
Figure 21. Single-wheel tester with parallel arms.
Most single-wheel testers use the mechanism shown in figure 22. Torque is measured at the input to the
wheel.
GT MR
slrrr
T
W
Va
ωω
Wd
With parallel arms, there is
no change in vertical reaction
as torque is applied (i.e., W = Wd)
NT
rt
GT MR
slrrr
T
W
Va
ωω
Wd
NT
rt
Figure 22. Parallel arm single-wheel tester with direct measurement of NT.
Using Tractors to Test Tires
While single-wheel testing is conceptually the most simple and direct approach to testing a traction
device, a tractor can be used to test tires (fig. 23) and in fact is the most common method to test belts. Use
of a tractor means that two of the traction devices are actually under test. When a tractor is used, some
method must be used to determine the dynamic weight on the drive wheels (due to weight transfer when
pulling), and in the case of full four-wheel drive, some method must be used to determine the net traction

22
(pull) developed by each axle. For a tractor with mechanical front-wheel drive, it is nearly impossible to
conduct a pure traction test because the dynamic weights and the net traction developed by the front axle
need to be determined, and the front and rear axles are typically equipped with different size tires.
c
MR
slrrr
T
RWS
ωω
GT
Va
FWS
MRf
RWD
Wb
P
Dh
FWD
rt
c
MR
slrrr
T
RWS
ωω
GT
Va
MRf
RWD
Wb
P
Dh
FWD
rt
Figure 23. Two-wheel drive tractor used as tire tester. Front static weight (FWS) is usually settobeapproximately20%of
total tractor weight.
When using a tractor to test tires, it is best to use a two-wheel drive model while measuring or
calculating the front axle dynamic weight and motion resistance force. Since there is no torque input at the
front axle, i.e., no gross traction force, the motion resistance force is equal to a negative net traction and is
subtracted from the measured drawbar pull for traction calculations. (Note: For force/moment calculations,
the total drawbar pull should be used.)
When using a complete tractor to test tires or belts, certain measurements are needed or calculated
before testing can begin. Front (FWS) and rear (RWS) static weights must be adjusted and measured. It is
helpful if the static front weight (FWS) is kept relatively low (similar to that required for implements
attached to the drawbar) and at a constant percent of the total tractor weight. Use of front and rear suitcase-
style weights (fig. 23) facilitates obtaining the correct weight on the test tires (rear) and the desired weight
distribution. Drawbar height (Dh) and wheelbase (Wb) are measured for use in calculating weight transfer
to the rear and rear dynamic weight (RWD). The rolling radius (rr) of the tire under test is determined by a
"free roll", that is, zero drawbar pull and zero net traction, as described by method 2 (in section 2 of this
paper under "Travel Reduction Ratio"). If the free roll is determined on a hard surface, then the front
motion resistance (MRf) is ignored.
During a test, the input axle torque (T) and speed of the axle (ω) are measured along with the drawbar
pull (P) and actual travel speed (Va). The rear dynamic weight (RWD) is calculated from the rear static
weight (RWS) and the weight transfer from the front. In the calculations, an attempt is made to calculate
(estimate) the front motion resistance force (MRf), which is the difference between drawbar pull (P) and
the net traction (NT) output from the tires under test. In addition, recognize that there are actually at least
two tires under test.
The rear dynamic weight (RWD) of the tractor is calculated as the sum of the static rear weight (RWS)
and the weight transfer. Weight transfer is calculated as P × Dh/Wb. Since the drawbar pull is horizontal,
any weight transferred to the rear axle is subtracted from the front, and the front dynamic weight (FWD) is
calculated by subtracting the weight transfer from the front static weight (FWS). This is then used to
estimate the front motion resistance. The method used to calculate front motion resistance (MRf) is a
circular reference in a spreadsheet. Assume that the front motion resistance ratio is equal to that calculated
at the rear. This calculation can also be done on-the-go in a data acquisition program. The steps taken are:
1. Initially assume that the dynamic front weight (FWD) is equal to the static front weight (FWS).

23
2. Assume a front motion resistance ratio (MRR) as a starting point.
3. Calculate front motion resistance force (MRf) and add it to the net traction force (initially equal to
drawbar pull).
4. Recalculate the new rear MRR (from within the traction calculations using new NTR).
5. Use the newly calculated rear MRR to calculate the new front (MRf).
6. Loop back until the data do not change (to whatever significance you are operating, usually the third
decimal place). At this point, the front MRR is equal to the rear MRR, and the net traction force has
been adjusted for what it is estimated to be required to push the front wheels through the field. Note
that this means that net traction will be higher than the measured drawbar pull (considering the number
of tires under test).
Traction testing in the field always requires a way to apply a load to the tire test tractor. It is possible to
take traditional drawbar loading units that might be used on a hard surface to the field and use them.
However, there are problems in doing this. Relatively heavy load units are required for the high draft loads,
which limits the minimum drawbar pull to the motion resistance of the load unit, and as a result, lower
portions of the pull-travel reduction curve cannot be attained. It is preferred to use a second live tractor in
combination with an implement (fig. 24). An implement that operates narrow and deep is preferred, to limit
the area of the test field that is disturbed. The travel speed of the load tractor (gear and engine speed, set at
about 3/4 throttle to allow adjustment up and down) is set to match that of the test tractor. The implement is
operated at a depth that the load tractor can pull, and the drawbar pull applied to the test tractor is adjusted
using the throttle of the load tractor. Throttling back the load tractor increases the load on the test tractor, as
the test tractor is thereby caused to pull more of the implement load. Increasing the throttle setting on the
load tractor causes it to overtake the test tractor, reducing the drawbar pull as the load tractor pulls the
implement. The disadvantage of this method is that it requires two operators and a method to communicate
between the two as the load is adjusted.
c
RWS
Va
FWS
RWD FWD
Tire Test Tractor
Vt
Dh
c
Load Tractor
Dh
cc
Wb
Figure 24. Tire test tractor with powered tractor and implement as load unit.
Speed Effects
The speed or gear of the test tractor is set to one that can develop full torque (desired wheel slip) to the
test tires. Other methods can be used, depending on the flexibility of the transmission of the tire test tractor,
but it is necessary to go from the minimum torque necessary to self-propel the tractor to that which
develops in the range of 30% to 40% travel reduction. Depending on the power of the tire test tractor and
its weight (the weight on the traction device being tested), this torque could be developed over a range of
travel speeds. This leads to the question of speed effects in traction testing.
Figure 25 shows the results of a traction test conducted over an extreme (for agricultural tractors) range
of travel speeds. The test was conducted under self-propelled conditions, i.e., net traction is zero (except for
possible front axle effects). The test showed a slight increase in MRR with speed, but little change over the
normal range of tractor field speeds. It is easy to be misled by the extra power required to drive the tractor
at the higher speeds, as the axle power ranged from about 5 to 35 kW while the MRR was virtually
constant.

24
Theoretical (no slip) Travel Speed (Vt, m/s)
MotionResistanceRatio
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.0 1.0 2.0 3.0 4.0 5.0 6.0
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
AxlePower(kW)
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0.0 1.0 2.0 3.0 4.0 5.0 6.0
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
Axle Power, (kW)
4.5 m/s = 10 mph
MRR = 0.0079(Vt) + 0.0806
Figure 25. Speed effects on motion resistance. Tire = 710/70R38 on loose soil.
5. Traction Performance
The following series of figures shows typical traction test curves for different soils, tire size and
pressure, load on the tires, and belts vs. tires. All curves were developed using the regression methodology
on data from actual traction tests. Figure 26 is an example that shows how to interpret the traction plots. As
indicated in figure 26, each tractive efficiency curve has a peak value, a value of net traction ratio where the
peak occurs, and an arbitrary efficient pull (net traction) range. Also shown is the resulting travel reduction,
including maximum pull (net traction ratio), which is usually limited by the test engineer. Tractive
efficiency tends to peak near an NTR of 0.4 for tires and 0.5 for belts.
TractionPerformanceRatios
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Efficient Pull Range
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Tractive Efficiency Ratio
Peak TE
Pull @ Max T E
Max Pull
(slip limited)
Figure 26. How to interpret traction plots (performance of 20.8R42 dual tires on three surfaces).
Effects of Soil
Figure 27 shows the performance of 20.8R42 dual tires on three tractive surfaces. Note that the peak
tractive efficiency is reduced as the soil becomes softer and looser, but that the peak occurs at a net traction
ratio of approximately 0.4 on all soils. Maximum NTR is also reduced as the soil becomes less firm (lower
net traction at the same travel reduction).

25
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
UntilledTilled
Subsoiled
Tractive Efficiency
Figure 27. Performance of 20.8R42 dual tires on three surfaces (8300 kg axle load, 83 kPa tire pressure).
Effects of Tire Pressure
Figure 28 is an example of operating tires that were over-inflated for their load capacities compared to
correctly inflated tires. Maximum tractive efficiency is reduced as well as the maximum net traction ratio.
While the maximum tractive efficiency is only reduced about 5 percentage points, the effect on vehicle
performance may be significantly greater depending on how implements are matched to the tractor. For
example, at an NTR of 0.5, more than 10 percentage points of difference will result in about 17%
difference in drawbar power and significantly higher travel reduction.
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Net Traction Ration (NTR)
Tractive Efficiency
6 psi load @ 6 psi (correct)
(41 kPa) (41 kPa)
6 psi load @ 28 psi
(41 kPa) (193 kPa)
Figure 28. Performance of single tire (Firestone 710/70R38 ATR) at two inflation pressures in tilled (loose) tractive
conditions.
Effects of Tire Size
Performance of two tire sizes at the correct inflation pressure is shown in figure 29. The larger diameter
520/85R46 tire shows significantly higher power efficiency at the same slip. This demonstrates that wider
tires (710/70R38) have greater motion resistance loss (both operating at about the same travel reduction).
Both provide similar maximum pull capability, which may give the operator the impression that there is no
performance difference between the tires.

26
1.00
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90
Tractive Efficiency
710/70R38, 6 psi load @ 6 psi (correct)
(41kPa) (41kPa)520/85R46, 8 psi load @ 8 psi (correct)
(55kPa) (55kPa)
Figure 29. Performance of two sizes of single tires at correct inflation pressures in tilled (loose) tractive conditions.
Effects of Load on Tire
Figure 30 shows performance of the same size tire with correct inflation pressure for different weights.
Using the correct inflation pressure for the weight allows the tire to operate at its design deflection ratio
where optimum performance is obtained (Zoz and Turner, 1994). Maximum tractive efficiency is achieved
at about the same net traction ratio for each weight. It should be noted that this means pull in proportion to
the dynamic weight. In other words, if only the weight is changed, then performance may suffer from not
operating at the optimum NTR.
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Tractive Efficiency
(110kPa) (110kPa)
(55kPa) (55kPa)
Figure 30. Performance of single tire (Goodyear 520/85R46 DTR) at two weights with correctpressuresintilled(loose)
tractive conditions.
Belt and Tire Comparisons
To determine the effect of belt width on tractive performance, three belts of 400, 630, and 810 mm
width (16, 25, and 32 in.) were tested from one manufacturer. A 400 mm (16 in.) belt was tested from a
second manufacturer. During these tests, a set of 20.8R42 dual tires was also compared. Tests were
conducted on untilled and tilled tractive conditions. In the process of testing, a subsoiler was used on the
load tractor to provide a portion of the load. This tilled pass provided a soft subsoiled condition on which
further tractive comparisons were made.

27
Figures 31 through 33 show the performance of the three belt widths in comparison to the dual tires on
three surfaces. Under firm untilled conditions (fig. 31), there is little performance difference between the
four treatments at normal field pulls (NTR approximately 0.4 to 0.5). Dual tires dropped off at the higher
pulls, and the wider belt provides higher maximum NTR (limited by travel reduction).
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
TravelReductionRatio
Cat25”(630mm)
20.8R42 Dual Tires
Tractive Efficiency
Cat 32”(810mm)
Cat 16”(400mm)
Figure 31. Belt width comparison on firm untilled soil (belted tractor total weight = 12700 kg; wheel tractor axle weight =
8303 kg).
As the tractive conditions become softer and looser, the differences are mo re evident between belts and
tires, while the belts maintain their relative position with each other (figs. 32 and 33). Note that the
maximum tractive efficiency for tires still comes at NTR of about 0.4, while the belts tend to maximize at a
slightly higher pull (approximately 0.5 NTR) and demonstrate a wider range of pulls at near-maximum
tractive efficiency. It should be noted that these tests were all carried out with a minimum of steering.
Matching implements at higher VTR on the skid steer belted machine may hinder steering control under
load.
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
20.8R42 Duals
Cat 16”(400mm)
Cat 32”(810mm)
Tractive Efficiency
Cat 25”(630mm)
Figure 32. Belt width comparison on tilled soil (belted tractor total weight = 12700 kg; wheel tractor axle weight = 8303
kg).

28
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
NetTractionRatio(NTR)
TractiveEfficiency
Cat32”(810mm)
Cat 25”(630mm)
Cat16”(400mm)
20.8R42Duals
TravelReductionRatio
Figure 33. Belt width comparison on subsoiled land (belted tractor total weight = 12700 kg; wheel tractor axle weight =
8303 kg).
Two manufacturers' 400 mm (16 in.) belts are compared in figures 34 through 36. Little difference was
observed in performance between the two belts on the three surfaces tested. It should also be noted that
under the firm untilled conditions, the correctly inflated 20.8R42 dual tires equaled or outperformed the
400 mm (16 in.) belts.
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Tractive Efficiency
Cat 16”(400mm)
20.8R42 Duals
GY 16”(400mm)
Figure 34. Belt manufacturer comparison on firm untilled soil (belted tractor total weight = 12700 kg; wheel tractor axle
weight = 8303 kg).

29
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Tractive Efficiency
20.8R42 Duals
GY 16”(400mm)
Cat 16”(400mm)
Figure 35. Belt manufacturer comparison on tilled soil (belted tractor total weight = 12700 kg; wheel tractor axle weight =
8303 kg).
1.00
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90
20.8R42 Duals
Tractive Efficiency
Cat 16”(400mm)
GY 16”(400mm)
Figure 36. Belt manufacturer comparison in subsoiled tractive condition (belted tractor total weight = 12700 kg; wheel
tractor axle weight = 8303 kg).
Figures 37 and 38, respectively, compare the performance of 630 mm (25 in.) belts and 20.8R42 dual
tires on the three surfaces tested. As the soil becomes softer, dual tires demonstrate greater difference in
tractive efficiency than does the belt.

30
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Untilled
Subsoiled
Tractive Efficiency
Tilled
Figure 37. Performance of 630 mm (25 in.) belt on three surfaces (belted tractor total weight = 12700 kg; wheel tractor
axle weight = 8303 kg).
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
1.00
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
UntilledTilled
Subsoiled
Tractive Efficiency
Figure 38. Performance of 20.8R42 dual tires on three surfaces (axle load = 8300 kg; tire pressure = 83 kPa).
6. Soil, Tire, and Traction Equations
Tractive performance is affected by both the soils' normal strength and its shear strength. In general,
normal strength has the most effect on motion resistance, while shear strength has the most effect on travel
reduction. Describing and documenting the soil is perhaps the most difficult part of traction testing. There
are several reasons for the difficulty. First, the soil has sufficient variation, which can easily influence the
soil sampling device. Second, soil measurements are time consuming, and finally, the sampling technique
may not be replicated or repeated for different soil conditions. For this reason, much of the traction tests
reported are of a comparative nature, that is, one traction device compared to another device while operated
under the same soil conditions. Relating the traction conditions from one set of tests to another is not easy
because describing the soil itself is not easy.
The device that is the most portable and commonly used, the cone penetrometer (fig. 39), works well
only if the soil has moisture and if it has not been disturbed. Soil strength as measured by the soil cone
penetrometer provides a combined measurement of soil normal strength and shear strength. The cone
penetrometer also requires a large number of measurements because there is a large variability in the test
results. Cone penetrometer (Perumpral, 1987) testing involves pushing a cone of a specific size and shape

31
into the soil at a certain rate and recording the resisting force exerted by the soil on the penetrometer (ASAE
Standards, 2001c).
Figure 39. ASAE standard cone penetrometer for soil.
The force required to push the cone into the ground is recorded as a function of depth. The force
divided by the area of the base of the cone provides a "pressure" measurement and is referred to as cone
index. Common units are lb/in2
in English units and kN/m2
(megapascal) in SI units. Cone index may be
measured as deep as 500 mm (20 in.) when used for tillage and/or compaction measurements, but the upper
100 to 150 mm (4 to 6 in.) is commonly used for traction purposes. The engineering practice (ASAE
Standards, 2001a) provides a systemic approach for observing and presenting cone index data.
If tractive performance is to be predicted, then mathematical expressions for the interaction of the
traction device and the soil must be developed. Traction equations have been developed and reported by
several researchers (Wismer and Luth, 1972; Brixius, 1987; Upadhyaya et al., 1988; Zoz and Brixius,
1979), but the most commonly accepted are those reported by Brixius (1987). The equations were
developed to predict the tractive performance of bias-ply tires operating in cohesive frictional soils. Tire
torque, motion resistance, net traction, and tractive efficiency are predicted as a function of soil strength,
tire load, travel reduction (slip), tire size, and tire deflection. As bias-ply tires are no longer in common use,
modifications proposed by Brixius (1987) and verified by Al-Hamad et al. (1994) for radial-ply tires are
now generally used.
The Brixius (1987) equation uses three relationships:
Mobility number:




















+






+






=
d
b
31
h
ä
51
W
CIbd
Bn (14)
Torque ratio:
( )( ) 0.04e1e10.88
rW
Q
GTR 7.5S0.1Bn
+−−== −−
(15)
Motion resistance ratio:

32
Bn
0.5S
0.04
Bn
1.0
W
M
MRR ++== (16)
Refer to table 1 and figure 40 for definitions of the symbols in equations 14 through 16
Table 1. Parameters in the Brixius equation.
Symbol Parameter (and Dimension)
Soil
CI = cone index (FL-2
)
Wheel
b = unloaded tire section width (L)
d = unloaded tire diameter (L)
r = tire rolling radius (L)
ä = tire deflection (L)
h = tire section height (L)
System
W = vertical wheel load (F)
S = slip
Q = wheel torque (FL)
M = motion resistance (F)
P = wheel pull (F)
Figure 40. Tire cross-section (from Brixius, 1987).
The Brixius traction equations predict the GTR and the MRR as a function of the mobility number and
slip (travel reduction). The mobility number is a dimensionless combination of soil and tire parameters. It is
used in the traction equations to predict the combined effect of the soil and tire parameters. Curve-fitting
techniques were used by Brixius (1987) to determine the manner in which the dimensionless ratios
combine.
It should again be noted that the Brixius (1987) equations as presented above apply to bias-ply tires.
Table 2 compares the original coefficients for bias-ply tires to those presently in use. Changes were made
to more accurately represent the results from recent tests on radial-ply tires.
Table 2. Brixius traction equation coefficients for bias-ply and radial-ply tires.
Original coefficients
for bias-ply tires
Coefficients proposed
by Brixius
for radial-ply tires
Coefficients currently
used in tractor
performance spreadsheet
7.5 8.5 to 10.5 7.0
0.88 0.88 0.88
0.1 0.1 0.08
0.04 0.03 to 0.035 0.03
1 0.9 1.20

33
Calculating tractive performance using the Brixius (1987) equations requires an iterative solution and is
best handled in a computer program or spreadsheet. Calculating tractor performance adds another iterative
dimension, as the dynamic weight on the tires depends on the pull (due to weight transfer), and the pull
(calculated from the traction equations) depends on the weight. A spreadsheet has been developed to
calculate tractor performance and is described in section 7 of this paper.
7. Tractor Performance
Tractor performance is not the same as tractive performance. Tractor performance is proportional to the
performance of the traction device(s), but not equal to it. The primary difference is that tractive
performance (efficiency) depends on knowing the input power (axle power) to the traction device. Axle
power for a complete tractor is seldom known and is not measured during official tests. But there are also
other reasons that tractive performance and tractor performance differ:
• Tractive performance is given for a defined tire or traction device, for example an 18.4R46 tire. A
tractor may operate with a combination of different traction devices, that is, different size tires on
the front and rear axles.
• Due to weight transfer when operating, even if a tractor has the same tires front and rear (4WD
tractors, for example), both the static and dynamic weight that the tires are operating with will
likely be different between the front and rear axles, requiring different tire pressures and thus a
"different" traction device.
The performance of a tractor depends on the performance of a combination of traction devices and the
performance of the tractor drivetrain (table 3).
Table 3. Comparison of traction device and tractor.
Traction device Tractor
Axle power input Axle power usually unknown
Tractive efficiency (from axle to
"drawbar")
Power delivery efficiency (from PTO or engine to drawbar)
Net traction (NT) Drawbar pull
Net traction ratio, NTR Vehicle traction ratio (VTR)
Travel reduction ratio Travel reduction ratio (but may differ at different drive wheels)
Motion resistance ratio (MRR) Requires tow test to measure
Gross traction ratio Not applicable; impossible to calculate
Dynamic weight on device Variable between axles
Single tire size May have multiple sizes
Single tire pressure (at given load) May have multiple pressures
Tractor size defined by PTO or engine power
While the efficiency of a traction device is defined as tractive efficiency, the efficiency of a complete
tractor is defined as power delivery efficiency (Zoz et al., 2002). Power delivery efficiency (PDE) is
defined as the ratio of the delivered drawbar power of a tractor to the vehicle input power of the tractor. It
represents the percentage of power produced by the engine that is available as tractive power delivered
through the drawbar (Shell et al., 1997; Turner et al., 1997, Zoz et al., 2002). Tractive efficiency itself is
defined as the ratio of output power to input power of a traction device (ASAE Standards, 2001b). PDE
includes TE and the efficiencies of the entire vehicle drivetrain from the engine to the drawbar.
PDE is computed by dividing drawbar power by a specified input power measured at some location
behind the engine. Exactly where this input power is measured may vary with the specific vehicle. It is
advantageous for the input power to be measured at a location that defines the power level of the tractor
being tested. For many tractors, this is the power take-off (PTO). For vehicles where size is commonly
specified by engine power, such as 4WD tractors or tractors without PTOs, engine flywheel power may be
the only power measurement used. When engine flywheel power or axle power is used as the input, it is
normally measured directly during tractor comparison testing.
When PTO power is used, it is commonly predicted during tractor comparison testing using laboratory-
derived regressions from previously correlated engine parameters such as engine speed, fuel rack position,
and injector needle lift duration. While power measured at different locations such as engine flywheel,

34
PTO, transmission output, or drive axles can be used in the PDE calculation, if two tractors are to be
correctly compared to one another, then the power used for the PDE calculation must be measured at the
same point on each tractor. It is also most meaningful if the location used is that which defines the size of
the tractor, which is the advertised power.
PDE is most accurate as a comparator if the engine flywheel power can actually be measured and used
in the calculation. Measured engine flywheel power is used directly in the PDE calculation as follows:
PowerEngine
PowerDrawbar
PDE = (17)
When it is necessary to use regression analysis or other method to determine the equivalent PTO
performance, the following equation is used:
PowerPTOEquivalent
PowerDrawbar
PDE = (18)
Power delivery efficiency depends on the performance of the PTO and transmission drivetrains as well
as the performance of the tires or other traction device. Various power losses can be summarized as:
PTO power = Engine power - PTO drivetrain loss - Hydraulic loss from PTO operation.
Axle power = Engine power - Transmission drivetrain loss - Hydraulic loss from transmission
operation.
Drawbar power = Axle power × tractive efficiency.
and
EfficiencyDrivePTO
EfficiencyTractiveEfficiencyDriveonTransmissi
PDE
×
= (19)
It becomes obvious that tractor performance depends on more than just the performance of the traction
device(s).
Figure 41 is an example of tractor power delivery efficiency comparing a belted and rubber-tired tractor
of similar design. Data for tractive efficiency (more precisely, axle power delivery efficiency) is shown in
figure 42.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Vehicle Traction Ratio, VTR
TractorPerformanceRatios
PTO POWER DELIVERY EFFICIENCY
TRAVEL REDUCTION RATIO
TIRE VEHICLE
BELTED VEHICLE
BELTED VEHICLE
TIRE VEHICLE
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
TractorPerformanceRatios
PTO POWER DELIVERY EFFICIENCY
TIRE VEHICLE
BELTED VEHICLE
BELTED VEHICLE
TIRE VEHICLE
Figure 41. Belted and tire vehicle power delivery performance in primary tillage using PTO power calculated from engine
power.

35
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
TIRE
BELT TRACTIVE EFFICIENCY
BELT
TIRE
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
TIRE
BELT TRACTIVE EFFICIENCY
BELT
TIRE
Figure 42. Belt and wheel tractive performance in primary tillage.
The difference between tractive efficiency and power delivery efficiency is due to drivetrain
efficiencies. Figure 43 shows approximate efficiency relationships for an agricultural tractor drivetrain
(wheeled tractor).
GROSS FLYWHEEL
NET FLYWHEEL
TRANSMISSION INPUT
AXLE PTO0.96
.85-.90 .90-.92
.77-.80
.91
.92
.82-.84
.89-.91.84-.88
.99
GROSS FLYWHEEL
NET FLYWHEEL
TRANSMISSION INPUT
AXLE PTO0.96
.85-.90 .90-.92
.77-.80
.91
.92
.82-.84
.89-.91.84-.88
.99
Figure 43. Approximate agricultural tractor power relationships (drivetrain).
Since the concrete test track is pretty much a constant for all tractors under official tests, the tests
become primarily a test of drivetrain performance. The difference in tire performance from concrete to field
conditions becomes the biggest variable in tractor field performance.
Tractor Performance Spreadsheet
Traction equations for agricultural tires were developed by Brixius (1987) from extensive tire tests and
were described in section 6 of this paper. The original equations were for bias-ply tires, which are no longer
in common use, especially on higher-powered tractors. Radial-ply tire coefficients were suggested by
Brixius (1987) and verified by Al-Hamad et al. (1994). The equations allow us to calculate expected tractor
performance given tractor physical specifications, including tractor axle power. The calculations require an
iterative solution and are best accomplished with a computer or in a spreadsheet. A series of Lotus 123
speadsheets was developed by Zoz (1987). These templates could be used to predict 2WD and

36
4WD/MFWD tractor performance on agricultural soils. Drawbar performance including pull, speed, power,
and wheel slip were calculated. An alternative calculation mode was provided to determine the required
static weight necessary to obtain a given performance, i.e., a tractor weight guide. The spreadsheet was
later updated to Excel and has been widely distributed in the ASAE community (fig. 44).
Figure 44. Excel spreadsheet for predicting tractor field performance under field conditions.
The spreadsheet has two modes of use, either performance calculation or tractor weight calculation;
inputs and outputs change accordingly.
Common inputs (for performance calculation and weight calculation):
• Rear tire size. A data table is provided for the most common R1 tires of radial-ply or bias-ply
construction. The initial value taken is that used for Nebraska Tractor Testing Laboratory (NTTL)
tests when tractor is selected.
• Front tire size. A data table is provided for the most common R1 radial-ply and bias-ply tires and
F2 tires for 2WD tractors. The initial value taken is that used for NTTL tests when tractor is
selected.
• Number of tires. Allows use of single, dual, or triple tires. The initial value taken is that used for
NTTL tests when tractor is selected.
• Tractor wheelbase. Taken from NTTL data when tractor is selected.
• Hitch height. Initially taken as the drawbar height from NTTL data when tractor is selected.
However, it must be changed for other hitch configurations.
• Draft angle. The initial draft angle is zero as taken from NTTL data. However, it can be varied for
drawbar operation and must be changed along with hitch height and rear location for use with
hitch-mounted implements.
• Hitch location behind rear axle. This data is not given by NTTL and is not necessary for
operation with horizontal drawbar pull. The spreadsheet automatically inserts nominal values (that
can be changed) when implement type is selected.

37
• Input power. Axle power is used by the spreadsheet. However, as stated earlier, axle power is
seldom known for a tractor, even from official tests. The spreadsheet uses an input power and a
transmission efficiency to calculate axle power. A tractor database is provided which takes data
from NTTL tests. The efficiency of power transmission from PTO or engine to axle has been
calculated from NTTL data by back calculation, assuming tire efficiency on concrete of 92.5%.
This value was taken from extensive traction tests of tires on concrete. It results from
approximately 2% motion resistance and 5% travel reduction at maximum tractive efficiency.
Input power is split between front and rear on 4WD/MFWD tractors by the dynamic weight and
wheel mobility number.
• Travel speed, theoretical (Vt). The initial value is 5 mph, which probably should be changed.
The theoretical speed is the speed without travel reduction and is the speed that is normally given
in tractor specifications for a tire of a given size. If a tractor is equipped with radar, it is the speed
that would be measured without slip (travel reduction) at the appropriate engine speed.
• Soil strength. Soil strength is given by the soil cone index. Figure 45 lists approximate values for
good, medium, and poor tractive conditions.
Figure 45. Typical cone index values (from Brixius, 1987).
Additional inputs for performance calculation (see fig. 46):
• Tractor static front axle weight. Initially taken as unballasted NTTL test weight, this value
should be adjusted.
• Tractor static rear axle weight. Initially taken as unballasted NTTL test weight, this value
should be adjusted.

38
Resultant Implement
force
c
Va
FWS
RWS
RWD
Wb
FWD
HT+
DIS+
ang+
Tractor Performance Prediction
Spreadsheet
Vt
Dh
Pull
Inputs:
Traction Equations
Tire size, type, no
Radial, Bias, Singles
Duals, Triples
Soil Strength, Cone Index
Db Power
Wheel Slip
Power split
Tire pressures
P or U
DIS+
Outputs
Axle Power
Resultant Implement
force
Resultant Implement
force
Resultant Implement
force
c
VaVa
FWSFWS
RWSRWS
RWDRWD
WbWb
FWDFWD
HT+HT+
DIS+DIS+
ang+ang+
Tractor Performance Prediction
Spreadsheet
VtVt
DhDh
PullPull
Inputs:
Traction Equations
Tire size, type, no
Radial, Bias, Singles
Duals, Triples
Soil Strength, Cone Index
Db Power
Wheel Slip
Power split
Tire pressures
P or U
DIS+DIS+
Outputs
Axle PowerAxle Power
Figure 46. Tractor diagram for tractor performance prediction spreadsheet (performance calculation).
Additional inputs for weight calculation:
• % Slip (travel reduction). Percent slip is selected; necessary front and rear static weights are
calculated.
• % Dynamic front weight desired. Front and rear static weights are calculated for given front
dynamic weight percentage and desired travel reduction.
Common outputs:
• Drawbar pull.
• Actual travel speed.
• Drawbar power.
• Correct tire pressures for given or calculated static weights (does not consider increases that may
be necessary for transport of heavy implements).
• Power delivery efficiency.
• Vehicle traction ratio.
• Tractor weight to power ratios.
Estimated Drawbar Power
As with estimates of drivetrain efficiencies, estimates of drawbar performance are possible. Figure 47
estimates the drawbar power for different types of tractors on different soil conditions.

39
0.96
CONC GOOD MED POOR TRACTOR CONC GOOD MED POOR
TYPE
0.91 0.75 0.70 0.57 0.87 0.72 0.67 0.55
0.91 0.79 0.75 0.66 MFWD 0.87 0.76 0.72 0.64
0.92 0.80 0.78 0.73 4WD 0.88 0.77 0.75 0.70
0.92 0.85 0.83 0.81 BELT 0.88 0.76 0.74 0.72
AXLE PTO
DRAWBAR
0.96
TYPE
Axle Power Delivery Efficiency PTO Power Delivery Efficiency
AXLE PTO
2WD
Figure 47. Approximate agricultural tractor power relationships (drawbar).
8. Optimizing Tractor Drawbar Performance
The major factors affecting tractor drawbar performance during field operations are tires and ballasting.
Tires are usually selected at the time of purchase, while ballasting changes can be made at any time. In
practice, ballast weights are not often changed as soil and operational conditions change.
Tires
From a tractor performance standpoint, the general rule for tire selection should be "bigger is better."
The second rule should be that larger diameter is preferred over larger width, again from a performance
standpoint. In either case, tires should be operated at the correct pressure for the load being carried. Lower
tire pressures are helpful from a compaction standpoint, as ground pressure is roughly equal to tire
pressure. In addition, lower pressures help control power hop as they allow a wider range of tire pressure
adjustments to be made. Larger tires can operate at lower pressures for the same weight. Correctly inflated
radial-ply tires provide a 5% to 7% efficiency improvement over bias-ply tires.
Ballasting
Ballast is weight added to the tractor for the purpose of improving the tractor's performance. Depending
on the field conditions and the drawbar requirements of the operation, the tractor's unballasted weight may
actually be heavier than the optimal weight. Agricultural tractor ballasting recommendations have evolved
over the years, based primarily on field experience (Bloome et al., 1983). In the early 1970s, John Deere
published a slide rule (OBM-20R2) to calculate recommended tractor weight based on available power and
speed of operation. In the 1980s, NIAE developed a similar formula (Dwyer, 1984; Gee-Clough et al.,
1982) relating weight, power, and speed of operation to size of tire and ballasted weight. Recent traction
tests and work with the revised Brixius (1987) traction equations (for radial-ply tires) have led to a better
understanding of the variables involved and provided a technical basis for recommendations that have been
in use for a number of years.
Developing ballasting criteria requires an understanding of the objective of proper ballasting. There is
probably not universal agreement, and the criteria may change to meet special situations, such as where
mobility and flotation may be the primary requirement. However, for most situations encountered by
agricultural tractors, the objective is to optimize the time spent during field operations at near maximum
power delivery efficiency in order to maximize the work completed and to minimize the fuel consumption.

40
While a net traction ratio of 0.4 (for tires) has been shown to be optimum for best tractive efficiency, it
is difficult to use as a ballasting criterion because the drawbar pull of a farmer's tractor is seldom known.
The gross traction ratio for maximum tractive efficiency varies from about 0.46 to 0.52 over a range of soil
conditions (fig. 48). Since tractors must operate over a range of travel speeds and since ballast weights are
not likely to be changed as tractor operations change, a compromise is needed. Tractors tend to be ballasted
for the most severe or heaviest load operations, and these levels are maintained for other operations. If
tractor PDE is to be optimized over a range of operations, then the tractor must be lighter than required for
full load operation in the worst condition.
Gross Traction Ratio (GTR)
0.30 0.50 0.70
1.0
0.9
0.8
0.7
0.6
0.5
0.4
4 6
8
10
20
40
30
TractiveEfficiencyRatio
40
25
15
10
BN
ConcreteTravel Reduction, %
>>
MAX TE GTR = 0.54
0.40 0.60 0.80
1.0
0.9
0.8
0.7
0.6
0.5
0.4
4 6
8
10
20
40
30
40
25
15
10
BN
ConcreteTravel Reduction, %
>>
MAX TE GTR = 0.54
Figure 48. Optimum gross traction ratio (Brixius equation revised for radial-ply tires).
Dynamic weight transfer from implements must be considered, and since the implement usually
transfers weight to the tractor, the static weights required are further reduced. A gross traction ratio in the
order of 0.54 to 0.60 provides a level of ballast that is practical for most operations.
The gross traction ratio for maximum tractive efficiency is nearly independent of soil strength and tire
size (Bn), and since it is defined by the input to the traction device, its value can be calculated from tractor
performance parameters of axle power and speed. Gee-Clough et al. (1982) and Dwyer (1984) both
developed expressions for tractor weight in the form of
Speed
Constant
Power
wheelsdrivingonWeight
= (20)
Dwyer (1984) stated that: "For a surprisingly wide range of tyres and soil conditions it has been found
that tractive efficiency reaches a maximum at a coefficient of traction of about 0.4." Assuming a tractive
efficiency of 70% and a coefficient of traction (net traction ratio) of 0.4, he found the following relationship
for weight, power, and speed:
175
0.4
0.7
Power
peed)(Weight)(S
== (21)
This can be rewritten as:
NTR
TE
Power
peed)(Weight)(S
= (22)

41
But since
GTR
S)-(NTR)(1
TE = (23)
and speed = actual travel speed = Va
and 





−=
Vt
Va
1S
Therefore
GTR
1
Power
t)(Weight)(V
= (24)
where Vt = theoretical travel speed.
This supports the use of gross traction ratio as the basis for ballasting recommendations. Note that
weight on driving wheels is dynamic weight (Wd), a combination of static weight and weight transfer from
implements or other axles. Speed should be the theoretical travel speed (Vt) for a given gear or
transmission speed setting (including tire size). Power is that expected from the axle of the tractor. Axle
power is seldom known or measured directly and must be estimated from engine or PTO power.
In its simplest form, WSP-1
= 1/GTR. The constants in the expression depend on the system of units
being used. In the form shown, it is dimensionless. However, customers seldom use purely SI units. Some
conversions are necessary, and a constant k can be applied, depending on the units to be used (see table 4):
GTR
WSP 1 k
=− (25)
Table 4. Logical values for k.
Units m/s mph km/h
kg/kW 106 -- 383
lb/hp -- 375 630
Using the English lb/hp-mph system and GTR = 0.54:
694
0.54
375
WSP 1 ==− (26)
or typical SI system application with kg/kW and speeds in km/hr:
709
0.54
383
WSP 1 ==− (27)
It is interesting to note that the value for WSP-1
is, for all practical purposes, the same for these
combinations of units (about 700). It is also necessary to keep in mind that the power is at the axle, weights
are dynamic, and speeds are without slip (travel reduction).
Figure 49 is a plot of the WSP-1
relationship for SI units with the suggested optimum GTR shown. Note
that slower speeds require heavier weights. Note also that the weight depends on the value assumed for
GTR, and is greater for the lower GTR values.

42
5 7 9 11 13 15
No-Slip Travel Speed (km/h)
150
140
130
120
110
100
90
80
70
60
50
0.55
0.50
0.60
TractorWeight,kg/axlekW
0.40
0.45
5 7 9 11 13 15
150
140
130
120
110
100
90
80
70
60
50
0.55
0.50
0.60
Gross Traction Ratio0.40
0.45
Optimum Gross Traction Ratio
Figure 49. Weight-speed-power relationship.
From this discussion, a gross traction ratio of 0.54 provides an optimum level of ballast to allow for
varying travel speeds, tractor (engine) loading, soils, and operation with different implements. This value is
plotted as the dashed line on the graphs of figures 48 and 49.
Figures 48 and 49 use axle power and theoretically apply to the tire only. It is still necessary to make
the step from "tire only" and "axle power" to a tractor and PTO or engine power. Some detailed differences
in tire performance can be expected between front and rear axles of tractors with either equal or unequal
tire sizes. Further differences are expected between driven and undriven axles (2WD). However, in the
overall picture, the total weight levels do not vary widely and are probably compensated for by changes in
the operating speed with 2WD (faster). When all axles are driven, the total tractor weight can be used
without much problem, as long as appropriate weight distributions are considered.
Converting fromaxle power to commonly used PTO or engine power specification requires applying a
simple multiplier to the WSP-1
number after the ratio of axle power to PTO or engine power is determined.
Typically:
Axle power = (0.96)(PTO power) (28)
PTO power = (0.85)(Flywheel power) (29)
Using these relationships and assuming GTR = 0.54 results in a table of values for different units of
measurements and locations where the power is specified (table 5). Note that the power should be what is
being used during field operation. It may need to be adjusted for tractor loading (i.e., % of full load).
Table 5. Ballast numbers (GTR = 0.54).
Units m/s mph km/h
Axle power
kg/kW 196 -- 709
lb/hp -- 694 1166
PTO power
kg/kW 190 -- 680
lb/hp -- 670 1120
Flywheel power
kg/kW 160 -- 580
lb/hp -- 570 952

43
The values in table 5 are equivalent to what Gee-Clough et al. (1982) referred to as the "ballast
number." Dividing these numbers by the speed gives necessary weight-to-power ratios. Alternatively,
dividing these numbers by the weight-to-power ratios gives the appropriate speed of operation. For
reference, the John Deere Tractor Ballasting Guide (slide rule) used 625 and 530, respectively, for PTO and
flywheel power. The following are examples of how to use the ballast numbers:
Tractor weight (lb/PTO hp) = Ballast number / No-slip travel speed
= 670 / 5 mph
= 125 lb/PTO hp
Theoretical travel speed (Vt, mph) = Ballast number / lb/PTO hp
= 670 / 130 lb/PTO hp
= 4.8 mph
As previously stated, the weights involved need to be considered dynamic weights, that is, including the
effects of weight transfer from implements and other axles. Weight is normally transferred to the tractor
from implements, and analysis will show that the amount transferred is a function only of the draft angle
(the remainder is transferred from one end of the tractor to the other and does not significantly affect the
overall results from a ballasting standpoint). Draft angles can vary widely, but for equipment hitched to the
drawbar, draft angles in the 5° to 10° range are not uncommon.
If a vehicle traction ratio of 0.4 is assumed as the expected result of proper ballasting, then calculated
tractor weights can be reduced by 0.4 times the tangent of the draft angle. This results in 4% to 8%
reduction in weight requirement for draft angles in the range of 5° to 10°.
Ballasting Sensitivity
After all the calculations are made regarding the tractor weight and amount of ballast, the question
becomes how accurate and precise the relationships need to be. As stated earlier, the objective of ballasting
is to optimize the time spent during field operations near maximum PDE over a span of time. Depending on
how often ballast change may be made, this may be weeks, years, or the life of the tractor.
In order to make an analysis of how critical ballasting really is, the Brixius (1987) traction equations
and Zoz (1987) spreadsheet were used to predict the optimum weight requirements for a John Deere 8870
tractor in 860 kPa (125 psi) cone index soil with 20.8R42 dual tires. Starting with 77 kg/axle kW
(approximately 105 lb/flywheel hp) tractor weight, the maximum PDE was found to be at about 9.8 km/hr
operating speed, close to the 0.54 GTR in figure 50. Using this as a starting point, a locus of points of 95%
and 97.5% of the maximum tractive efficiency were determined.

44
5 7 9 11 13 15
150
140
130
120
110
100
90
80
70
60
50
0.55
0.50
0.60
GTR
5 7 9 11 13 15
TheoreticalTravelSpeed(Vt,km/hr)
150
140
130
120
110
100
90
80
70
60
50
TractorWeight(kg/AxlekW)
0.40
0.45
5 7 9 11 13 15
150
140
130
120
110
100
90
80
70
60
50
0.55
0.50
0.60
GTR
5 7 9 11 13 15
150
140
130
120
110
100
90
80
70
60
50
0.40
Allpointsinsidecurvearewithin5%of
optimum efficiency at 77 kg/axle kw
Optimum efficiency
at 77 kg/axle kW
2.5% curve
Soil: 860 kPa (125 psi ) CI
Tire: 20.8R42 D
Figure 50. Optimum tractor weight and sensitivity analysis (John Deere 8870 4WD).
It is interesting and significant to note the wide range of either weight (to power ratio) or travel speed
that allows operation within the 2.5% and 5% contours. For example, tractors operating at 8 km/hr could
range from about 70 to 97 kg/kW and still be within 2.5% of the maximum efficiency for these conditions.
Or, more likely, the 77 kg/axle kW tractor could operate from about 7.5 to 13 km/hr and still be within
97.5% of the maximum efficiency. The operator would likely notice a significant change in wheel slip over
this speed range, but the efficiency would not change all that much.
For the analysis shown in figure 50, the tire inflation pressures were automatically adjusted for the
required weights, that is, higher tire pressures were used with heavier weights. Since lower pressures were
required at the higher speeds, this is the likely reason that the 2.5% and 5% contours are open to the right
and do not close up on themselves.
While ballasting is important to optimize a particular situation, bigger gains would probably be made at
the time of purchase by proper sizing of tires to allow operation at lower tire pressures (increasing Bn) at
any given weight.
Ballasting Limitations
Using the ballast numbers to calculate required weights to operate at a given speed assumes sufficient
tire capacity to operate within a reasonable range of tire inflation pressures, and pressures must go up as
speed is decreased and weight increases. At some point, the maximum pressure and load carrying capacity
for the tire will be reached. Of equal concern are the high loads and axle torques that may be applied to the
tractor's drive train at slow operating speeds; some tractor operating limits may be reached. Adding too
much ballast reduces drive train life and may cause unnecessary soil compaction due to use of higher tire
pressures.
Ballast Optimization in the Tractor Performance Spreadsheet
The tractor performance prediction spreadsheet provides an optimization routine to determine the
tractor weight and weight distribution for maximum drawbar power. The details of the optimization scheme
are discussed in detail in Jones and Grisso (1992). A macro uses a golden section search as the optimization
algorithm when the spreadsheet is in the weight calculation mode. The algorithm reviews the results of
changing the weight distribution and travel reduction and searches for the optimal power delivery
efficiency.

45
9. Conclusions
This paper has reviewed traction mechanics and tractive performance parameters to predict the tractor
performance during field operations. By maximizing the tractive advantage of the tractor, more efficient
tractor operations can be achieved and the cost of operation can be minimized. The terms for testing
tractors and developing predictive equations have be reviewed and implemented into a computer model
(spreadsheet) so that operational characteristics can be accessed. Finally, ballast and tire criteria were
discussed. It is the hope of the authors that these discussions will provide insights and stimulus into the next
revolutionary step in the design and application of traction devices and tractor designs.
Toward that end, several items are noted for future investigation and consideration. The tractive
predictions for belted vehicles need to be characterized and built into a computer model with similar ballast
and belt operational criteria as has been developed for vehicles with tires. It is also recognized that steering
a belted vehicle with a load has certain limitations and that further development is needed. One concept that
has merit is the design of an elliptical wheel where the advantages of both of the traction devices (belt and
tire) can be embraced. The concept could change the footprint length depending on the operational needs.
An elliptical wheel could also make the tire footprint narrower to improve performance and match crop
production schemes using narrow rows.
Pressure for the development of new tractor concepts with new roles and tasks will have an impact on
field performance. For example, the development of tires and suspensions for high-speed tractors will need
to be assessed concerning their impact on field performance. It will be unlikely that the best tires for
transport will also be the best for fieldwork. The role of inflation systems and the development of control
systems for them have the potential to improve field productivity. The role of a robotic tractor and its
performance requirements will need to be developed and assessed. The role of a lighter tractor and concepts
to reduce soil compaction are still important.
The assessment of the soil conditions and the enhanced soil's spatial and temporal variations are
essential for understanding and predicting tractor operations. The understanding of soil responses from
vehicle traffic continues to need practical insights. Traction is attained from a complex interaction of
traction devices and soil conditions. Theory, experiment, and field tests have revealed the general nature of
this interaction and have provided some insights for analysis of present and future tractions systems.
However, plenty is left to investigate and understand before a comprehensive traction mechanics is
established.
Acknowledgements
Due to the outstanding contribution to traction and tractor performance made by W. W. Brixius, the
authors would like to acknowledge that she changed her name to Christina M. Hollis in April 1992. Dr.
Hollis retired from Deere after 29 years in December 2001. Since then, she has continued to be contracted
through Deere for various hydraulic, power train, and cooling modeling projects. Presently she owns a
company called Omega Consulting, Inc., located near Elizabeth, Illinois.
10. References
Al-Hamad, S. A., R. D. Grisso, F. M. Zoz, and K. Von Bargen. 1994. Tractor performance spreadsheet for
radial tires. Computers and Electronics in Agric. 10(1): 45-62.
ASAE Standards. 2001a. EP542. Procedures for using and reporting data obtained with the soil cone
penetrometer. St. Joseph, Mich.: ASAE.
_____. 2001b. S296.4. General terminology for traction of agricultural tractors, self-propelled implements,
and traction and transport devices. St. Joseph, Mich.: ASAE.
_____. 2001c. S313.3. Soil cone penetrometer. St. Joseph, Mich.: ASAE.
Bloome, P. D., J. D. Summers, A. Khalilian, and D. G. Batchelder. 1983. Ballasting recommendations for
two-wheel and four-wheel drive tractors. ASAE Paper No. 831067. St. Joseph, Mich.: ASAE.

46
Brixius, W. W. 1987. Traction prediction equations for bias-ply tires. ASAE Paper No. 871622. St. Joseph,
Mich.: ASAE.
Brixius, W. W., and R. D. Wismer. 1978. The role of slip in traction. ASAE Paper No. 781538. St. Joseph,
Mich.: ASAE.
Corcoran, P. T., and D. S. Gove. 1985. Understanding the mechanics of track traction. In Proc. Int'l
Conference on Soil Dynamics, 4: 664-678. 17-19 June. Auburn, Ala.: Auburn University, Office
of Continuing Education.
Dwyer, M. J. 1984. The tractive performance of wheeled vehicles. J. Terramechanics 21(1): 19-34.
Gee-Clough, D., M. McAllister, and G. Pearson. 1982. Ballasting wheeled tractors to achieve maximum
power output in frictional-cohesive soils. J. Agric. Eng. Res. 27(1): 1-19.
Jones, D., and R. D. Grisso. 1992. Golden section search as an optimization tool for spreadsheets.
Computers and Electronics in Agric. 7(4): 323-335.
Perumpral, J. V. 1987. Cone penetrometer applications – A review. Trans. ASAE 30(4): 939-944.
Shell, L. R., F. M.Zoz, and R. L. Turner. 1997. Field performance of rubber belt and MFWD tractors in
Texas soils. In Belt and Tire Traction in Agricultural Vehicles, 65-73. SAE SP-1291. Warrendale,
Pa.: Society of Automotive Engineers.
Turner, R. J., L. R. Shell, and F. M. Zoz. 1997. Field performance of rubber belt and MFWD tractors in
southern Alberta soils. In Belt and Tire Traction in Agricultural Vehicles, 75-85. SAE SP-1291.
Warrendale, Pa.: Society of Automotive Engineers.
Upadhyaya, S. K., W. J. Chancellor, D. Wulfsohn, and J. L. Glancey. 1988. Sources of variability in
traction data. J. Terramechanics 25(4): 249-272.
Wismer, R. D., and H. J. Luth. 1972. Off-road traction prediction of wheeled vehicles. ASAE Paper No.
72619. St. Joseph, Mich.: ASAE.
Zoz, F. M. 1987. Predicting tractor field performance (updated). ASAE Paper No. 871623. St. Joseph,
Mich.: ASAE.
Zoz, F. M., and W. W. Brixius. 1979. Traction prediction for agricultural tires on concrete. ASAE Paper
No. 791046. St. Joseph, Mich.: ASAE.
Zoz, F. M., and R. L. Turner. 1994. Effect of correct pressure on tractive efficiency of radial-ply tires.
ASAE Paper No. 941051. St. Joseph, Mich.: ASAE.
Zoz, F. M., R. L. Turner, and L. R. Shell. 2002. Power delivery efficiency: A valid measure of belt and tire
tractor performance. Trans. ASAE 45(3): 509-518.

The CD-ROM contains:
Traction performance spreadsheet
Estimated power spreadsheet
Lecture manuscript in PDF format
Lecture manuscript in MSWord format
Powerpoint of the visuals used in the presentation
Selected references

View publication statsView publication stats

Traction and tractor_performance

More Related Content

What's hot (20)

Similar to Traction and tractor_performance (15)

More from yared seifu (9)

Recently uploaded (20)

Traction and tractor_performance